[A] Even Sir Humphry Davy considered the attraction of the pole as
  being communicated from one particle to another of the same kind
  (483.).

543.  Neither does M. de la Rive’s theory, as I understand it, require that the particles should be in combination:  it does not even admit, where there are two sets of particles capable of combining with and passing by each other, that they do combine, but supposes that they travel as separate compounds of matter and electricity.  Yet in fact the free substance cannot travel, the combined one can.

544.  It is very difficult to find cases amongst solutions or fluids which shall illustrate this point, because of the difficulty of finding two fluids which shall conduct, shall not mingle, and in which an element evolved from one shall not find a combinable element in the other. Solutions of acids or alkalies will not answer, because they exist by virtue of an attraction; and increasing the solubility of a body in one direction, and diminishing it in the opposite, is just as good a reason for transfer, as modifying the affinity between the acids and alkalies themselves[A].  Nevertheless the case of sulphate of magnesia is in point (494. 495.), and shows that one element or principle only has no power of transference or of passing towards either pole.

  [A] See the note to (670.).—­Dec. 1838.

545.  Many of the metals, however, in their solid state, offer very fair instances of the kind required.  Thus, if a plate of platina be used as the positive pole in a solution of sulphuric acid, oxygen will pass towards it, and so will acid; but these are not substances having such chemical relation to the platina as, even under the favourable condition superinduced by the current (518. 524.), to combine with it; the platina therefore remains where it was first placed, and has no tendency to pass towards the negative pole.  But if a plate of iron, zinc or copper, be substituted for the platina, then the oxygen and acid can combine with these, and the metal immediately begins to travel (as an oxide) to the opposite pole, and is finally deposited there.  Or if, retaining the platina pole, a fused chloride, as of lead, zinc, silver, &c., be substituted for the sulphuric acid, then, as the platina finds an element it can combine with, it enters into union, acts as other elements do in cases of voltaic decomposition, is rapidly transferred across the melted matter, and expelled at the negative pole.

546.  I can see but little reason in the theories referring the electro-chemical decomposition to the attractions and repulsions of the poles, and I can perceive none in M. de la Rive’s theory, why the metal of the positive pole should not be transferred across the intervening conductor, and deposited at the negative pole, even when it cannot act chemically upon the element of the fluid surrounding it.  It cannot be referred to the attraction of cohesion